Last week I attended a wonderful event, the First Conference on Faculty Development in the Health Sciences. It covered a number of topics in educational research. Faculty development involves educating faculty in the roles for which they lack training. Educational research differs from laboratory science or even clinical medicine. It's messy. Randomized prospective trials are virtually impossible, and sample sizes tend to be small.
Mixed methods research can overcome some of these limitations by providing a richer data set for study. The workshop presentation showed how mixing quantitative methods (those that generate a p value after statistical analysis) with qualitative methods (those that identify the how or why of the data) can provide more information and make studies worthwhile (and publishable).
A post over on the Renal Fellow Network today sparked my memory. They posted a survey on their site to look at the perceived impact of limited resident duty hours. As residents work less, the work must still be done, often by fellows. The results somewhat surprised Matt Sparks, the fellow posting the survey:
Of 65 respondents, 43% felt they were working harder than in residency, while 33% felt the workload was similar. Sparks goes on to consider the reasons:
I can only speculate as to why these two different opinions exist. It could be secondary to the different types of nephrology training programs (research oriented vs. clinically focused). Or could be due to the difference in work load allocation (e.g. strong hospitalist presence vs. light hospitalist presence).
What if this study then could identify those responding in the two main groups and follow-up with structured interviews or written open-response surveys? More information about fellowship type, comparison of inpatient responsibilities, structure of inpatient services, and other attitudes and impressions could clarify the survey data and lead to a testable hypothesis.
I realize that the present survey cost essentially nothing; however, the fellows involved should consider taking the next step to make this into a revealing research project. And a great example of the utility of mixing methods.